ILS 2012

Québec, Canada, 26 — 29 August 2012

ILS 2012

Québec, Canada, 26 — 29 August 2012

Schedule Authors My Schedule

THEMATIC SESSION: Robust Supply Chain Management II

Aug 27, 2012 04:30 PM – 06:00 PM

Location: VCH-2860

Chaired by Keenan Yoho

4 Presentations

  • 04:30 PM - 04:52 PM

    Agility Approach Design and Foundation : Case of Lumber Industry

    • Dhia Eddine Boughzala, presenter, Université Laval
    • Mustapha Nourelfath, Université Laval
    • Jean-Marc Frayret, Polytechnique Montréal, CIRRELT
    • François Léger, Consortium de recherche FORAC, Université Laval

    This paper develops an agility approach in the lumber industry. The industrial context and the business environment are explored. An integrated approach is based on four dimensions: conceptual, contextual, experimental and decisional. The first two dimensions give the specification components of the agility as a wide-range vision and are based on lumber industry specificities. The experimental framework aims to validate the components’ correlation and the interaction between environment, performance and actions. The decisional dimension models the agility continuous improvement.

  • 04:52 PM - 05:14 PM

    Comparison between Deterministic and Stochastic Production Planning Approaches in Sawmills via Simulation

    • Naghmeh Vahidian, presenter, Concordia University
    • Masoumeh Kazemi, Concordia University
    • Mustapha Nourelfath, Université Laval

    The goal of this paper is to compare the performance of stochastic and deterministic production planning approaches in sawmills on a rolling planning horizon. A series of designed experiments are proposed to identify significant factors influencing the performance of production planning process in sawmills. As it is not possible and not economically reasonable to interrupt the production line in sawmills to test different production planning models, Monte-Carlo simulation is used for implementation of different plans. By reviewing and analyzing the simulation and experimental results, our goal is to propose a decision framework for guiding the decision maker to select among deterministic and stochastic production planning models under different circumstances. Two influencing factors are identified based on the experimental results for a realistic scale case study and a decision framework is proposed based on the analysis of the results. The decision framework differs for each KPI. If the backorder has the priority for decision making, deterministic model has better performance in lower demand levels. On the other hand, for higher demand levels the stochastic model performs better than the deterministic model in terms of backorder cost. If the inventory is defined as the important KPI, the impact of interaction between length of planning horizon and demand level should be considered as the basis of decision making. For short planning horizons, deterministic model has superior performance in higher demand levels than the stochastic one. In contrary, for long planning horizons, the deterministic model can be the appropriate choice for lower demand levels and stochastic model has better performance for higher demand levels.

  • 05:14 PM - 05:36 PM

    A Scenario Decomposition Algorithm for Stochastic Sawmill Production Planning with Set-up Constraints

    • Masoumeh Kazemi Zanjani, presenter, Concordia University
    • Daoud Aït-Kadi, Université Laval
    • Mustapha Nourelfath, Université Laval

    We study a real-world multi-period, multi-product production planning problem involving set-up constraints, with random yield and demand. The resulting large-scale multi-stage stochastic mixed-integer model cannot be solved by mixed-integer solvers of commercial optimization packages. The production planning model is a mixed-integer programming (MIP) model without any special structure. As a consequence, developing efficient decomposition and cutting plane algorithms to obtain a good solution in a reasonable amount of time is not straightforward. We propose a solution strategy based on the progressive hedging algorithm (PHA), which iteratively solves the scenarios separately. The proposed approach attempts to gradually steer the solutions of the sub-problems towards an implementable solution by adding some penalty terms in the objective function used when solving each scenario. The solution of this strategy is a local optimum and an upper bound for the optimal objective value of the multi-stage stochastic model. Computational experiments for a real world large-scale sawmill production planning model verify the effectiveness of the proposed solution strategy in finding quickly a good approximate solution.

  • 05:36 PM - 05:58 PM

    Information Technology for Uncertainty Management in Supply Chain

    • Jaymeen Shah, Texas State University
    • Francis A. Mendez, presenter, Texas State University
    • Hsun Ming Lee, Texas State University

    This manuscript explores the merge of modern database management technology, data modelling, and probability management for decision support for supply chain management. Data visibility and data integration are crucial concepts to a supply chain information system. These concepts together with adequate analysis tools provide decision makers what is necessary to manage the supply chain. This manuscript revisits information technology that has been proposed in the past, but that seems more feasible in the present. In particular, the use of simulations in concert with multidimensional databases, the analysis of the multidimensional cube, and their capabilities in support of uncertainty management in the enterprise with particular interest on supply chain management are explored.